4-2 THE PAINTED GLASS IN

lights of these two ^vindows is evident from the fact of their being the narrowest h'ghts in the building, and that the glass exactl}' fits them. Let us. then, re-arrange the glass upon this supposition, and put in No. 1 light what is now in No. 7 light ; in No. 2 what is now in No. 8 ; in No. 3 what is now in No. 9 ; in No. 4 what is now in No. 10 ; in No. 5 what is now in No. 11 ; in No. 6 what is now in No. 12 ; leaving the glass in Nos. 13, 14, 15, 16, 17, and 18 as it now is, and we shall find the Apostles arranged in a not uncommon order,^ and a perfect alternation preserved in the forms of the canopies, and in the colouring of the designs, throughout the upper tier of lights. Let us now put in No. 7 light the glass wdiich is in No. 20 light ; in No. 8 the remains of the canopy w^ork first mentioned in No. 1, and the portion of the crucifix in No. 1 ; in No. 9 wdiat is now in No. 4 ; in No. 1 1 the remains of the canop3'^-work first mentioned in No. 2, and the portion of the crucifix in No. 5 ; in No. 12 what is now in No. 23 ; in No. 20 the remahis of the canopy-work first mentioned in No, 5, and the portion of the crucifix in No. 3 ; and in No. 23 the remains of the canopy-work first mentioned in No. 3, and the portion of the crucifix in No. 6 ; leaving No. 10 blank, and the glass in Nos. 19, 21, 22, and 24 as it now is ; and we shall find, supposing the missing subject of No. 10 light to have been a duplicate of that in No. 19,^ and that the remains of the canop^^-w^ork, first mentioned in No. 6, belonged to it ; that not only will a perfect alternation in the forms of the canopies and the colouring of the subjects be preserved throughout the East windows, in the one, whether regarded in a horizontal or in a perpendicular direction ;^ in the other, when regarded in a horizontal direction — and it is obvious that a double alternation might, by a different arrangement, be produced in this as well as in the former window — but that the attitudes of

• It is possible that Nos. lo and 17 are same windows of the same building. 1

transposed. If St. Mathias were to take know of an instance as early as the latter tile place of St. Matthew, which there is part of the twelfth century, nothing in the order of the canopy design ' This alternation of design and colour is or colouring to prevent, the apostles would observable in many early Perpendicular bearranged as at Fairford Church, Glouces- windows. The following diagram may serve terKliire, with the single exception that to explain my meaning. Let the letters there St. .Matthew precedes St. Jude. arranged in a square, represent It is by no means an uncommon oc- A C four figures and canopies : and let currence to find in ancient glass, the same BD canopies A and B each have a hgures re[>eated in different, or even the red spire-ground, and blue niche